Objectives: To correlate prostate size estimates performed by single or multiple examiners through digital rectal examination (DRE) with volume measured by transrectal ultrasound (TRUS) and to propose measures for predicting prostate volume using DRE estimates in clinical settings.
Methods: Data from four sources were analyzed: (1) the Olmsted County community study of 397 patients examined by a single urology nurse, with TRUS measurements done by multiple examiners; (2) a community study in Stirling, Scotland, involving 480 patients with DRE and TRUS performed by one urologist; (3) baseline data from the Veterans Affairs Cooperative Study No. 359 in 1222 patients with DRE and TRUS measurements by multiple personnel at 31 centers; and (4) a clinical series of 100 men with DRE and TRUS by a single urologist.
Results: DRE estimates and TRUS volumes were significantly correlated (r = 0.4 to 0.9), but prostate size was underestimated by 25% to 55% for men with a prostate volume over 40 mL, depending on the study, with greater variability for studies involving multiple examiners. In one study that assessed prostate dimensions by DRE, posterior surface area (SA) correlated with overall TRUS volume (r = 0.4). According to receiver operating characteristic curves, SA showed a 70% and 76% chance of correctly identifying men with prostate volume greater than 30 or 40 mL, respectively; those with larger prostates were best distinguished by SA greater than 7 cm2 (sensitivity greater than 0.74, specificity greater than 0.50).
Conclusions: DRE underestimates prostate size, particularly if TRUS volume is greater than 30 mL. However, DRE estimates may help identify prostates likely to be larger than certain cutpoints by TRUS. Posterior SA may be useful as a preliminary assessment when prostate size is an important predictor of therapeutic response.